WILDLIFE SOFTWARE

                       Interactive computer program
                   for landscape-level habitat analysis
                                     
                    John L. Roseberry and Qingwang Hao
                                     
Authors' address during this research: Cooperative Wildlife Research Laboratory, Southern Illinois
University, Carbondale, IL 62901, USA. Current address for Qingwang Hao: AT&T Bell Labs, Room
3W-F07, 184 Liberty Corner Road, Warren, NJ 07059, USA.

Key words: Computer software, habitat analysis, habitat models, landscape metrics

     The ability to inventory, quantify, and evaluate habitat at various spatial scales is essential for
effective wildlife management. Traditional perspectives of habitat are expanding to address concerns
about biodiversity, fragmentation, and ecosystem management (Dunning et al. 1992). This shift in focus
from smaller to larger scales is concurrent with advances in remote sensing, Geographical Information
Systems (GIS), and methods for quantifying spatial attributes of landscapes. Land use and cover can now
be inventoried and classified over large geographic areas using satellite imagery (Lillesand and Kiefer
1987). Landscape composition and pattern can be described by a variety of metrics (O'Neill et al. 1988,
Turner 1990, Baker and Cai 1992, McGarigal and Marks 1994, Riitters et al. 1995), and habitat suitability
can be evaluated with various types of models (Verner et al. 1984). Realistically, however, many wildlife
managers still have limited access to these powerful tools because of restrictive requirements for
specialized computer hardware, software, and expertise.

     Our objective was to develop PC-based software that combined graphical, analytical, and
modeling capabilities into an easy-to-use program available to wildlife managers and resource planners
lacking more expensive and sophisticated GIS hardware and software. The result was an interactive,
Windows(tm) (Microsoft Corp., Redmond, Wash.) application program called the Habitat Analysis and
Modeling System (HAMS). This software enables users to graphically display, measure, modify, and
analyze composition and spatial patterns of digitized land cover maps and evaluate habitat suitability for
selected wildlife species or communities.

     HAMS accepts raster-format (grid cell) land use and cover maps of <12 classes. Depending upon
available memory, images as small as 17 x 17 and as large as approximately 2,400 x 2,400 rows and
columns can be displayed graphically. Images can be selected from existing files, or they can be created
using on-screen editing tools. Smaller study areas can be defined for detailed analysis. Displayed images
can be permanently or temporarily modified to simulate land use and cover changes or management
alternatives. Land cover types (classes) can be added, deleted, or combined. Changes can be made to
entire classes, or they can be limited to individual patches. Features can be added to or deleted from the
landscape using rectangle, polygon, or linear draw tools.

     Measurements of area can be taken for individual patches, entire classes, or user-drawn
rectangles or polygons. Linear distances between any 2 points on an image can be measured. Results are
available in English or metric units.

     Composition and spatial patterns of the original or modified landscape can be quantified using
selected metrics applicable to the entire landscape or individual classes. Landscape metrics currently
available include class proportions, contagion, dominance, diversity, class richness, relative class
richness, juxtaposition, interspersion, and shared edge among all combinations of classes. Patch
characteristics may be computed for individual classes or an entire landscape and include number of
individual patches, mean patch size, total perimeter, several nearest-neighbor metrics (McGarigal and
Marks 1994, Gustafson and Parker 1992), fractal dimension, and modified fractal dimension (Olsen et al.
1993). Certain patch metrics can be expressed in absolute terms or per unit area. Nearest-neighbor 
distances also can be computed for designated pairs of complimentary classes (e.g., woods-cropland). 
Discretionary buffer zones may be established around study areas to accommodate patches that extend 
beyond study-area boundaries and also to allow habitat patches in the immediate vicinity of the study 
area to be included in nearest-neighbor metrics. If desired, original pixels can be aggregated into larger 
pixels prior to calculation of patch metrics. Also, classes representing the landscape matrix Forman and Godron
(1986) can be designated temporarily as "null" and thus excluded from patch calculations.

     HAMS uses Pattern Recognition (PATREC) models (Williams et al. 1977) to evaluate habitat
suitability of original or modified landscapes. PATREC models were selected because: (1) their relatively
simple and uniform structure is amenable to on-line development using a graphical user interface (GUI),
(2) they require less-specific assumptions regarding functional relationships between habitat components
and species' needs and thus may be more appropriate for use with relatively coarse-grained habitat data,
and (3) they are sufficiently general to be useful at different spatial scales and levels of classification.
Species or community-specific models can be selected from existing files, or they can be created using
special dialog boxes. In addition to metrics computed from the displayed image, models can include
user-defined variables, thus permitting incorporation of explicit information not available from the classified 
image. Models created or edited during a session can be saved for future use. HAMS also can use
models created in a text editor and saved as an ASCII file.

     HAMS is written in the Borland (circle R) C++ 4.0 programming language (Borland (circle R) Int.,
Inc., Scotts Valley, Calif.). The program runs on IBM/AT-compatible microcomputers with i386 or above
processors and 28 MB RAM (216 recommended). A 14-inch color VGA monitor is required (17-inch color
SVGA recommended). Operating system requirements are MS-DOS Version 5.0 or later and
MS-Windows(tm) Version 3.1 or later.

     HAMS is available without charge from the Cooperative Wildlife Research Laboratory, Southern
Illinois University at Carbondale, Carbondale, IL 62901. The system is distributed on a single 3.5-inch
high-density diskette that includes a self-expanding install program and an example image file and
PATREC model. A user's guide-reference manual is provided. HAMS can also be obtained through the
Bird Monitor Bulletin Board (301-497-5831) or an anonymous FTP site at ftp.im.nbs.gov.

     Acknowledgments. Primary financial support for the development of HAMS was provided by the
Illinois Department of Natural Resources, Division of Wildlife Resources with funds from the Federal Aid in 
Wildlife Restoration Program (projects W-63-R(SI) and W-106-R). Supplemental funding was provided by the 
Cooperative Wildlife Research Laboratory. A. Woolf kindly reviewed and critiqued the program and manual. We 
especially wish to acknowledge the important contributions of B. J. Richards to an earlier
prototype of HAMS (Richards and Roseberry 1993).

                             Literature cited

BAKER, W. L., AND Y. CAI. 1992. The r.le programs for multiscale analysis of landscape structure using the GRASS 
geographical information system. Landscape Ecol. 7:291-302.

DUNNING, J. B., B. J. DANIELSON, AND H. R. PULLIAM. 1992. Ecological processes that affect populations in complex 
landscapes. Oikos 65:169-175.

FORMAN, R. T. T. AND M. GODRON. 1986. Landscape ecology. John Wiley and Sons, New York, N.Y. 619pp.

GUSTAFSON, E. J., AND G. R. PARKER. 1992. Relationships between landcover proportion and indices of landscape 
spatial pattern. Landscape Ecol. 7: 101 - 110.

LILLESAND, T. M., AND R. W. KIEFER. 198'. Remote sensing and image interpretation. John Wiley and Sons, New York, 
N.Y. 721pp.

MCGARIGAL, K., AND B. J. MARKS. 1994. FRAGSTATS--spatial pattern analysis program for quantifying landscape structure. 
Version 2.0. For. Sci. Dep., Oregon State Univ., Corvallis. 141pp.

OLSEN, E. R., R. D. RAMSEY, AND D. S. WINN. 1993. A modified fractal dimension as a measure of landscape diversity. 
Photogrammetric Eng. and Remote Sensing 59:1517-1520.

O'NEILL, R. V., J. R. KRUMMEL., R. H. GARDNER, G. SUGIHARA, B. JACKSON, D. L. EANGELIS, B. T. MILNE, M. G. TURNER, 
B. ZYGMUNT, S. W. CHRISTENSEN, V. H. DALE, AND R. L. GRAHAM. 1988. Indices of landscape pattern. Landscape Ecol. 
1:153- 162.

RICHARDS, B., AND J. L. ROSEBERRY. 1993. An interactive computer program for display, manipulation, and analysis of 
habitat data. Page 185 in K. E. Church and T. V. Dailey, eds. Quail  III: national quail symposium. Kansas Dep. Wildl. 
and Parks, Pratt.

RIITTERS, K. H., R. V. O'NEILL., C. T. HUNSAKER, J. D. WICKHAM, D. H. YANKEE, S. P. TIMMINS, K. B. JONES, AND B. L. 
JACKSON. 1995. A factor analysis of landscape pattern and structure metrics. Landscape Ecol. 10:23-39.

TURNER, M. G. 1990. Spatial and temporal analysis of landscape patterns. Landscape Ecol. 4:21 -30.

VERNER, J., M. L. MORRISON, AND C. J. RALPH, EDITORS. 1984. Wildlife 2000--modeling habitat relationships of 
terrestrial vertebrates. Univ. Wisconsin Press, Madison. 470pp.

WILLIAMS, G. L., K. R. RUSSELL, AND W. K. SEITZ. 1977. Pattern recognition as a tool in the ecological analysis of 
habitat. Pages 521-531 in Classification, inventory, and analysis of fish and wildlife habitat: proceedings of a 
national symposium. Phoenix, Ariz. U.S. Fish and Wildl. Serv. Biol. Serv. FWS/OBS78/76.

Software Editor. Smith